Date of Award

11-5-2024

Document Type

Thesis

School

School of Computing

Programme

Ph.D.-Doctoral of Philosophy

First Advisor

Prof.V.S.Shankar Sriram

Keywords

Cloud Service Selection, Fuzzy Sets, MCDM, IVIFS, Random Forest Classifier

Abstract

During the past few decades, cloud computing became a primary driver for the next generation of digital technology due to the increase in organizational performance and profitability based on a ‘pay-as-you-use’ fashion at anytime and anywhere across the globe. Cloud computing enables various enterprises to access pooled resources (like storage, network bandwidth, software applications, processing power, etc.) over the Internet with minimal Information Technology (IT) infrastructure and capital expenditure.

Indeed, the enormous popularity of cloud computing in both academia & industry over the decade has resulted in a wide range of similar cloud services offered by numerous service providers. Even though, cloud computing is a powerful service model, researchers, organizations, and academicians are still reluctant to utilize cloud services because of its uncertain and dynamic nature in terms of availability, security, and resource elasticity.

Cloud services are often associated with uncertainty in user feedback, levels of Quality of Service (QoS), availability zones and resources, etc. Since the cloud computing environment is prominent to dynamic nature, it creates a negative impact on the quality of cloud services. The uncertainty in the context of cloud services still remains a challenging research problem to identify trustworthy cloud service providers.

Further, the Inconsistent Service Ranking (ISR) and Rank Reversal Phenomenon (RRP) are the two major concerns that lead to an inefficient cloud service selection. To address these major concerns (i.e., uncertainty, ISR, and RRP), the thesis focusses on the development of hybrid Multi-Criteria Decision-Making (MCDM) techniques to identify trustworthy service providers in cloud computing environments. The major contributions of the thesis are highlighted as,

1. The performance of cloud services that evaluate based on the subjective assessment data (user feedback) and objective assessment data (real-time monitored QoS values) are imprecise and inconsistent. To address this, an Interval-Valued Intuitionistic Fuzzy Set (IVIFS) with a hybrid weight method is integrated with an MCDM technique for the identification of trustworthy service providers in the user feedback dataset

2. To address the uncertainty and develop an accurate trust prediction model for objective assessment data, a Picture Fuzzy Set (PFS)-based MCDM approach with Naïve Bayes is designed to predict the trust values of cloud services

3. An objective weighted scheme-based MCDM with enhanced accuracy normalization technique is presented to address the Inconsistent Service Ranking (ISR) and Rank Reversal Phenomenon (RRP) issues in cloud computing environments.

The different techniques that are formulated for research work are validated against three real-world datasets namely, (1) CloudArmor-a trust feedback dataset with a set of storage service providers, (2) Quality of Web Services (QWS)-real-time monitored QoS values of cloud services that ensure the quality of cloud services, and (3) CloudHarmony-provide metrics to analyze the performance of cloud services and the service status of different CSPs. To ensure the robustness of the proposed hybrid MCDM techniques, the results are compared with the state-of-the-art MCDM approaches and sensitivity analysis.

Download

Share

COinS